THE PHYSIOLOGICAL MEANING OF THE MAXIMAL OXYGEN INTAKE TEST

(1)

THE PHYSIOLOGICAL MEANING OF THE

MAXIMAL OXYGEN INTAKE TEST

Jere H. Mitchell, … , Brian J. Sproule, Carleton B. Chapman

J Clin Invest. 1958;37(4):538-547. https://doi.org/10.1172/JCI103636.

Research Article

Find the latest version:

(2)

THE PHYSIOLOGICAL MEANING OF THE MAXIMAL OXYGEN INTAKE TEST1

ByJERE H. MITCHELL,2 BRIAN J. SPROULE,3 AND CARLETON B. CHAPMAN

(From the Cardiopulmonary Laboratory of the Department of Internal Medicine, University of Texas Southwestern Medical School, Dallas, Texas)

(Submitted for publication October 10, 1957; accepted December 2, 1957)

According to Hill, maximal oxygen intake is

reached when oxygen intake per unit time has

at-tained ". . . its maximum and remains constant

. . . owingto the limitation of the circulatory and

respiratory systems (1) ." In more definitive terms, when one subjects a normal individual to progressively increasing workloads, allowing

suffi-cient time for recovery between each increment of work, a linear relation between workload and oxygen intake is found. Ultimately, maximal oxygen intake per unit of time is reached;

be-yond this point the workload can usually be in-creasedstill further but, ordinarily, oxygen intake levels off or declines.

The maximal oxygen intake which a normal individual can achieve is sometimes taken as an index to maximal cardiovascular function provided

pulmonary function is normal. If this concept is sound, the test may come to be of enormous value in the critical evaluation of normal and abnormal

cardiovascular function. The superiority of a test that measures cardiovascular capacity, over one that determines whether or not an individual canattain a single arbitrarily chosen workload, is apparent. A value for cardiovascular capacity should characterize a given individual precisely;

use of asingle workload merely places him in one of two groups. The difficulty, insofar as maximal

oxygen intake is concerned, is simply that its

physiological meaning is imperfectly understood. The view that cardiac capacity is the determinant

of maximal oxygen intake is surmise, not estab-lished fact. Ventilatory factors, for example, are usually said not to be involved in determining

maximal oxygen intake but have not really been ruled out. For that matter, the various studies

1 This work has supported by grants from the United States Public Health Service [H-2113 (C)], and the Dallas Heart Association.

2_Cardiac _trainee, National Heart Institute.

3Research Fellow.

that demonstrate some degreeof arterial

desatura-tion during heavy work are often thought to sug-gest that a pulmonary factor, either ventilatory

or diffusive, may be involved. Finally, the

inter-play between cardiac output and AV oxygen

dif-ference during heavy workneeds further scrutiny. Therecan be no doubt, theoretically, that the de-gree to which AV oxygen difference can

expand

is, like cardiac output, a determinant of maximal oxygen intake.

It is the purpose of this

presentation

to

ex-amine some of these

relationships by

direct

meas-urement in the hope that the

physiological

mean-ing of the maximal oxygen intake test can be

clarified.

METHOD

In order to establish normal data for this laboratory,

a maximal oxygen intake test similar to that described by Taylor,Buskirk, and Henschel (2) was applied to 65 normal men. A motor-driven treadmill was used for the exercise. For normal subjects, a 10 minute warmup period (3 mph at 10 per cent grade) was followed by

a 10 minuterestperiod. Thefirsttest run was then car-ried out, usually at 6 mph and zero grade. Before the treadmill was started, the subject was connected to a Rudolph two-way breathing valve by means of a rubber mouthpiece, his nose being completely closed with a nasal clamp. For the firstminute of the run, expired air was used to wash out a Douglas bag connected to the expiratory side of the valve. Expired air for analysis

was collected for the last minute of a two and one-half minute run. The use of this time period was based on the data of Taylor, Buskirk, and Henschel (2), Robin-son (3), and Donald, Bishop, Cumming, and Wade (4).

Analysis of the expired air was immediately carried out

using a Beckman E-2 magnetic oxygen analyzer and a

Liston-Becker infrared carbon dioxide analyzer. Du-plicate analyses were carried out with a Scholander ap-paratus at regular intervals. The volume of expired air was measured in a Tissot gasometer and was corrected to STPD. These data permitted calculation of oxygen consumption either by using change in nitrogen content for correction of expired (to inspired) volume, or by use ofanomogram (5). Pulse rate during exercise was measured electrocardiographically. Respiratory rate was

(3)

THE PHSYIOLOGICAL MEANING OF MAXIMAL OXYGEN INTAKE

obtained by use of a strain gauge assembly connected to

the expiratory side of the Rudolph valve. After a 10 minute rest period, the workload was increasedby raising the grade to 2.5 per cent, the speed being held at 6 mph, and the procedure repeated. Subsequently, the subject performed work atincreasing levels, the load be-ingincreasedeach time byraising thegrade2.5 per cent, until oxygen intake per minute leveled off. In most

sub-jects, the entire test required about an hour and a half. Ordinarily, the point at which the oxygen intake curve (plotted against workload) ceased to rise was taken as maximal oxygen intake. In mostinstances (72per cent), when the workload was increased beyondthat producing maximal oxygen intake, the value either remained un-changedor declined. In some cases, however,a relatively slight rise occurred, necessitating the formulation of strict criteria for deciding whether or not maximal, or near maximal, intake actually had been reached. Plots of oxygen intake against workload for the entire ma-terial (Figure1) showedthat, until a maximalvalue was attained, oxygen intake rose 142 ± 44 ml. with each in-crease in workload. If the rise was less than 142 minus 88 (twice the standard deviation), or 54 ml., the final value was accepted as the maximal intake, the assumption being that the subject had attained his true maximal value or had reached the beginning of a plateau and could not increase his intake very much more.

With most younger normal subjects, it was found to be unnecessarily time-consuming to use 2.5 per cent grade increases at the lower levels, since most of them did not attain maximal intake until grades of 10 per cent or more were reached. It was also noted that values for oxygen intake sometimes varied erratically at lower workloads, especially in anxious individuals or in those given to hyperventilation. In some young normal

sub-jects, maximal oxygen intake was not reached even at the highest grade possible (14.75 per cent), making it necessary to increase the speed as well as the grade. The highest workload required was 9 mph at 14.75 per cent grade. With older subjects, no such problem arose. Infact, in some instances slower speeds were used if the 6 mph setting proved to be too troublesome for the subject.

In order to test the repeatability of the method as per-formed in our laboratory, first and second trials were done on 15 normal subj ects and the data analyzed by standard statistical techniques (6).

The method for determining maximal oxygen intake was essentially unchanged when it was combined with determination of cardiac output and blood studies (15 normal subjects). Subjects for the more elaborate pro-cedures were chosen at random. Maximal oxygen intake was first determined by the usual method. At a later date both cardiac output and maximal oxygen intake were determined at rest and at three workloads. The dye dilution techniquewasemployedforestimatingcardiac outputs. Prior to beginning the tests, a PE 90 catheter was inserted through a thin-walled 16 gauge needle into the left antecubital vein and was passed proximally to the upper brachial. Then a similar, but shorter, catheter

was introduced into the brachial arteryof the same arm,

also _through athin-walled 16 gauge needle. Finally, the same method was used to introduce a long PE 90 catheter retrogradeinto the femoral vein to themidthigh

level. Local anesthesia was routinely used before

in-serting the 16 gauge needles. The catheters were at-tached to three-way stopcocks and were kept filled with normal saline solution containing a small amount of

heparin. For the actual measurement of cardiac output, 10 mg.of Evansblue (T-1824) was delivered at the end of the catheter in the brachial vein about one and one-halfminutes after the beginning of exercise (during the collection of expired air), it having been shown by

Donald, Bishop, Cumming, and Wade (4) that cardiac output and AV oxygen difference reach a steady state

after the first minute of exercise. Arterial blood samples

werecollected at one second intervals in tubes containing powdered heparin, using either a manual collection de-vice or an electrically operated collector. During the final 20 to 30 seconds of the exercise, blood samples

were collected anaerobically from all three catheters

(brachial artery, brachial and femoral veins) for

meas-urement of oxygen tension and content, carbon dioxide

content and pH. Oxygen tension was determined

polarographically (7), oxygen content and carbon di-oxide content by the Van Slyke and Neill method, and

pH by use of a Cambridge Research model pH meter.

Carbon dioxide tension was calculated froma Singer and

Hastingsnomogram (8).

Analysis of blood for dye content was carried out ac-cording to a previously published method (9), using a

Beckman model DU spectrophotometer. The data were

plotted and extrapolated by the method of Hamilton,

*I 11 III IV V VI VII

Work Load

FIG. 1. RELATION BETWEEN SUCCESSIVE INCREMENTS IN

WORKLOAD ANDOXYGEN INTAKE IN NORMAL MEN Heavy black line represents mean values at each load. Dashed lines represent standard deviation about each point.

(4)

JERE H. MITCHELL, BRIAN J. SPROULE, AND CARLETON B. CHAPMAN

TABLEI

Maximaloxygenintake in liters per

minute-Duplicate

_trials

in 15 normalmen

Trial

Subject I II

A. G. 2.94 2.89

J. M. 3.86 3.77

C. D. 3.46 3.34

J. D. 2.84 2.60

J.B. 2.99 2.93

B.A. 3.34 3.39

S.W. 2.45 2.40

C. M. 3.20 3.57

C. J. 3.51 3.65

L.C. 3.46 3.23

C. McR. 2.95 3.08

W. M. 2.99 3.27

J. W. 2.77 2.52

F.H. 3.13 3.14

N. H. 2.07 2.24

Mean 3.06 3.07

S.D. 0.46 0.44

Moore, Kinsman, and Spurling (10). Mean circulation time, "central" blood volume, and cardiac output were

then determined, also by the method of Hamilton and co-workers.

RESULTS

Comparisons of gas analyses done by the gas

analyzers, on the one hand, and the _Scholander

method (11), on the other, showed the former to

be highly reliable. Mean values for oxygen

con-tent in 17 gas samples chosen at random as the

work proceeded were 16.38 0.84 per cent for the gas analyzers and 16.42 0.90 per cent for

the

Scholander

method. Corresponding values

for carbon dioxide were 4.86 0.78 and 4.85 + 0.76 per cent, respectively. The polarographic

technique,

as

employed (7),

has been_shownto be

highly

reliable at all levels of oxygen tension. More

specifically,

the method yields values over

physiologic

venous and arterial ranges that

cor-respond closely

to

_pO2

_{values calculated} _from

si-multaneous Van

Slyke

analyses and from deter-minations

_by

the

microbubble

technique of Riley,

Proemmel,

and Franke (12). In very low

ranges

(3

to 10mm. Hg) the error isabout 5per centif valuesarecalculated from the

_{polarographic}

equation (7)

insteadof

being

read offacalibration curve. From 20 to 100 mm. Hg, the percentage

error

drops

toabout 2.5 per centand is still lower

at

_higher

oxygen tensions. Repeatability of the method at

relatively

low oxygen tensions can be

judged

from firstand second trials done on

eight

blood

samples

from different sources,

equilibrated

with knowngas mixtures

containing

6.1 and 12.6

percent oxygen

(42

and 89mm.

Hg

partial

pres-sure), respectively.

Thefirst trialsonthe mixture

containing

6.1 per cent oxygen

yielded

a mean

value of 43.2 + _0.9 mm.

Hg;

the mean for the

second trials was 42.9 + 0.8 mm.

Hg (n

=8 in

both

trials).

For the second

mixture,

the

cor-responding

mean values were 88.7 + 1.3 and

89.4 2.0mm.

Hg.

The

repeatability

of the maximal oxygen intake test, if

rigid

criteria for

_determining

the

_point

at which the maximal value has been attained are

ILEII

Oxygen intakes in normalmenofvarious agesatincreasingworkloads, includingthatproducingmaximaloxygenintake

6mph,

0grade Premaximal Maximal Postmaximal

Age

group L.1min. ml./Kg./min. L./min. ml./Kg./min. L./min. ml./Kg./min. L.Imin. ml./Kg.fmin

Mean 2.59 34.6 3.15 41.8 3.37 44.7 3.18 42.7

20-29 S.D. 0.27 2.7 0.49 4.8 0.51 3.9 0.42 _4.3

n 33 35 36 28

Mean 2.54 32.7 2.88 37.3 3.04 39.3 2.91 36.6

30-39 S.D. 0.31 3.2 0.38 3.3 0.39 3.3 0.54 _4.6

n 18 18 18 18

Mean 2.51 30.4 2.79 33.6 2.94 35.4 2.77 32.9

40-49 S.D. 0.44 2.9 0.59 4.3 0.60 3.3 0.46 _3.5

n 8 8 8 8

Mean 2.13 27.7 2.46 32.0 2.26 29.4

Over S.D.

50 n 3 3 3

(5)

541

TABLE III

Ventilation and oxygen removalratesin normalmen atincreasing workloads

6mph,

0grade Premaximal Maximal Postmaximal

Agegroup A* B* A B A B A B

Mean 61.3 43.0 81.7 39.2 94.9 36.0 95.2 34.0

20-29 S.D. 8.7 6.3 14.6 6.4 15.9 5.6 17.3 5.2

n 33 33 35 35 36 36 28 28

Mean 66.1 39.3 80.5 36.8 89.4 35.0 88.5 33.3

30-39 S.D. 13.3 6.4 14.9 6.8 17.3 5.7 18.8 5.7

n 18 18 18 18 18 18 12 12

Mean 66.3 38.2 81.0 35.2 88.8 32.7 - 82.7 34.3

40-49 S.D. 11.0 5.3 19.8 5.1 21.3 4.4 18.6 5.1

n 8 8 8 8 8 8 5 5

Mean 57.0 38.2 65.5 38.7 60.6 38.4

Over S. D.

50 n 3 3 3 3 3 3

*_A: _{Total ventilation} _{(expired volume)} _in _{liters per min.} _(BTPS). _B: 02 _removal_rate

(STPD)removed from each liter of air (BTPS) respired.

applied, is demonstrated in Table I, in which the

results of first and second trials, weeks to months

apart, areshown for 15 subjects. Variance

analy-sis discloses that almost all the variance is

at-tributable to interindividual differences; the

ran-dom component is quite small and the

intraindi-vidual component virtually negligible.

The mean values for maximal oxygen intake

obtained in normal men aged 20 and over are

giveninTableII.

In Table III are presented the mean values for

ventilation and oxygen removal rates in normal

men at increasing workloads including the

maxi-mal and postmaximal levels.

The values for oxygen intake, cardiac output,

AV oxygen difference and other hemodynamic

variables at restand atthe maximaloxygen intake

loadare shown inTableIV.

Table V presents mean values for arterial and venous oxygen content, oxygen tension, per cent

oxygen saturation, and pH, both at rest and at

maximal oxygen intake.

The values for cardiac output and AV oxygen

difference at the maximal oxygen intake and at a

higher workloadare presentedin Table VI.

DISCUSSION

The need foraneffective measure of circulatory

capacity is undisputed. The maximal oxygen

in-take test, tojudgefrom theviews of Hill (1) and

expressedas ml. of 02

of Bainbridge (13), may fill the need but its

de-terminants have not been sufficiently clarified to

enable one to accept the test as a measure of

cir-culatory capacity with assurance. The term

maximal oxygen intake must, itself, be viewed

skeptically. It seems clear from the work of Christensen and Ho6gberg (14) and Taylor, Bus-kirk, and Henschel (2) that the actual value

reached for a given individual depends on the nature of the physical activity. It is, therefore, maximal relative to a given set of conditions, which must be carefully defined, rather than in an ab-solute sense.

For one thing, the actual technique used is far from standard. In the present work, the tech-nique used by Taylor was followed fairly closely except that the subject was required to grip a

sup-porting shelf with his left hand while he ran. Another difference was the time interval between each workload. In our own work, 10 to 15 min-utes is allowed; Taylor, however, runs only one

workload per day and requires several days to

complete afull determination. Buskirk and Tay-lor's figure for an 18 to 29 year group of sedentary students and soldiers was 3.44 + 0.46 liters per minute

(15).

For our 20 to 29 year group the

(6)

TABLE IV

Cardiac output, stroke volume, calculated A VO,differences,

and other items in 15 normal subjects at rest and at maximal0sintake *

Standing Maximal 02 resting intake level

02 intake

(L./min.) 0.340 :1- 0.04 3.22 i 0.46

n 5 15

Cardiac output

(L./min.) 5.4 :1 0.8 23.4 ± 5.5

n 14 15

Pulserate 91 z1 17 187 1 10

n 14 13

Stroke volume

(ml.) 624 18 125 + 25

n 13 13

AV02 diff.

(ml./100 ml.) 6.5 0.7 14.3 i:2.5

n 5 15

"Central" volume

(liters) 1.71 ± 0.32 3.49 0.83

n 14 15

Appearancetime

(seconds) 11.9 4 2.2 5.9 1.3

n 14 15

Mean circ. time

(seconds) 18.5 i 3.1 9.4 :1 2.0

n 14 15

*_{Mean maximal 02 intake determined} samegroup: 3.15 0.31L./min.

previously on

20 to 29 (4.15 + 0.36 liters per minute) (16),

and that of Slonim, Gillespie,and Harold (4.05

0.39) (17) for 65 young trained men, were

pre-sumably determined to some extent by the high

degree of physical training their subjects had

reached (18). Buskirk and Taylor's comparable

figure for a group of young trained subjects was

3.95 + _0.43.

In order to render the test as reliable as

pos-sible, rigid criteria for determining the point at

which maximal oxygen intake is reached must be

established. If this is done, the test is quite

re-peatable (Table I) and a given value

character-izes an individual for several months at least.

The effect of age on the maximaloxygenintake

can be seen in Table II. At the initial exercise

level the oxygen intake is the same

for

the first

three age groups but the maximal intake values

for all fourage groups showa decrease with age.

Theseresults arecomparableto those of Robinson (3). It is clear, therefore, that the age trend

must be taken into account if the test is used for

functional evaluationof human subjects.

In Table III it is noted that ventilation reaches

its highest value at the maximal oxygen intake

level. There is no significant difference in the

ventilation at the maximal and the postmaximal level in a given age group. With increasing age

the level of ventilationatmaximalworkloads tends

to decline. Oxygen removal rates decline with

increasingworkloads in the first three age groups.

In the oldest group, the oxygen removal rate

ap-TABLE V

Mean valuesfor 0, content, 0,tension, percentsaturation, pHand hematocritfromonearterial and

twovenoussitesat restandduringexercise at maximaloxygenintake level

02content 02 tension Per cent Hct.

Site State ml./100_ml. mm.Hg sat. pH Percent

Resting 18.1 1.4 87 9 97.1 d 3.1 7.40 at 0.04 46.5 3.8

Brachial n 14 15 9 15 14

artery Max.02 19.3 :+: 1.9 88 =1 12 94.7 :1: 2.6 7.19 0.09 48.3 4 _3.4

n 14 15 9 15 14

Resting 9.3 =1 2.4 31 i 7 52.6 := 12.9 7.364 _0.03 _43.8_{41 2.7}

Brachial n 8 8 8 8 8

vein Max.02 4.5 =i 1.2 29 =1 8 24.9 =1: 7.8 7.13 0.04 48.3 1 3.0

n 6 6 6 6 6

Resting 6.6 d 2.5 30 + 11 37.9 d 16.2 7.35 d 0.07 45.3 d 2.6

Femoral n 8 8 8 8 7

vein Max.02 3.9 d 1.3 30:+1 7 19.9 1 6.8 7.11 0.05 48.2 d 3.2

n 8 8 8 7 8

(7)

TABLE VI

Valuesfor cardiac output and A V 0,difference in six normal subjectsatthe maximal oxygen intake andatahigherworkload

Maximal oxygen intake level Beyond maximal oxygenintake level

02 Cardiac AV02 02 Cardiac AV02

intake output diff. intake output diff.

J. B. 2.56 20.6 12.4 2.42 15.6 15.5

L.C. 3.23 25.9 12.5 3.03 21.1 14.4

A.G. 2.89 15.6 18.5 2.90 14.7 19.7

F. H. 3.01 23.2 13.0 2.96 17.1 17.3

C. Mc. 3.08 26.3 11.7 3.05 25.8 11.8

J.W. 2.47 14.1 17.5 2.52 15.1 16.7

Mean 2.87 21.0 14.3 2.81 18.2 15.9

S.D. 0.31 4.7 2.7 0.34 4.1 2.5

pears to remain constant with increasing work-loads.

The primary purpose of the study was not, however, merely to re-examine previously estab-lished characteristics of maximal oxygen intake but, rather, to inquire into its physiological mean-ing. It is obvious that, in terms of the Fick equa-tion, the main circulatory determinants must be cardiac output on the one hand, and AV oxygen difference on the other (19). The relative role played by each of these factors in determining the maximal oxygen intakeis seen in Table IV. Oxy-gen intake increased 9.5 times from the resting

state to the workload producing the maximal fig-ure; the corresponding figure for cardiac output is 4.3. Proportionately, pulse rate and stroke volume increased equally-about two times. At

the time the cardiac output reached its maximal level, the "central" blood volume, as calculated by the Hamilton method, had doubled. The ap-pearance time ofthe dye and the mean circulation time had decreased byone-half.

On the face of all this, one might assume with

Bainbridge (13) that cardiac capacity is the pri-mary determinant of maximal oxygen intake. Such an assumption, however, ignores the fact that the AV oxygen difference makes a contri-bution of its own to oxygen intake. Simple

in-crease incardiac output by 4.3 times, AV oxygen difference remainingat6.5 ml. per 100 ml., would increase oxygen intake from 340ml. (resting), to

no more than 1,500 ml. Widening of the AV oxygen difference by 2.2 times (from 6.5 to 14.3 ml. per 100 ml.), however, permits oxygen in-take to exceed 3 liters under the same conditions.

Were it not for the ability of the organism to

widen its AV oxygen difference, cardiac output would have to increase nearly 10 times to sup-ply 3,200 ml. of oxygen per minute tothe tissues. Ourmeanvalue for AV oxygen difference during heavy work is in close agreement with that found by Asmussen and Nielsen (20), and Freedman,

Snider, Brostoff, Kimelblot, and Katz (21), but higherthan that found by Christensen (22).

The magnitude of the maximal AV oxygen difference is determined by the maximal level

100

8)

101//

40 a im

0

20-i

ci~~~~~~~~~~~~~~~~~~~~~~_

0~~~~~~~~~ 2

FIG. 2. OXYGEN DISSOCIATION CURVES AT THREE LEVELS OF PH (33)

Insert shows A and B (open circles), representing values for per cent saturation corresponding to two

observed values for arterial PO2 (87mm. Hgat rest, and 88 mm. Hg during heavy work). A' and B' represent per cent saturation, calculated from Van Slyke determi-nations, for the same two pO2 values. The fall in per cent saturation is about 2 per cent in each case.

(8)

REST EXERCISE

FIG. 3. THE AV OXYGEN DIFFERENCE AT REST AND

DURING EXERCISE

Total height of each column represents oxygen

ca-pacity. Blackareas at topare unoxygenatedhemoglobin.

Shaded areas at bottom represent mixed venous oxygen

content.

reached by the oxygen content of arterial blood

on the one hand, and the lowest level reached by

the oxygen content of mixed venous blood on the

other. In Table V it is noted that there is a 1.2

ml.per 100ml. rise in arterialoxygencontentand

a slight fall in per cent saturation (about 2.4 per

cent) between rest and heavy exercise. These items are easily understood when it is noted that

the exercise also induces a fall in arterial pH, a

rise in hematocrit, and no significant change in

arterial oxygen tension (Figure 2). In any event, the small increases in arterial oxygen

con-tentinduced by exercise can contribute very little to the widening of the AV oxygen difference

(Figure 3).

The magnitude of the AV oxygen difference is

influenced most markedly byencroachment onthe

mixed venous oxygen content (23). Mixed ve-nous oxygen content (calculated from oxygen

intake, cardiac output, and arterial oxygen

con-tent) decreased from 11.6 ml. per 100ml. at rest to5.0ml.per 100ml. duringexerciseatthe

maxi-mal oxygen intake level. This encroachment on

themixedvenousoxygen content is madepossible by two mechanisms: One is the extent to which

blood perfusing working tissue can surrender its oxygen; the other is the shunting of blood away

from inactive areas as noted by Donald, Bishop,

and Wade (24). The lowest level to which the

oxygen content of blood flowing through

strenu-ously workingmuscle may fall is not known but it seems unlikely that it ever approaches zero in nor-mal subjects (25). An approach to these prob-lems was made in the present study by comparing oxygen content and tension of blood from brachial and femoral veins during running. The brachial

vein blood was drawn from the stationary arm. It is seen in Table V that heavy exercise reduced venous oxygen content to less than half its resting value in both samples, the femoral sample show-ing slightly greater reduction than the brachial. The oxygen content of the femoral venous sample

(3.9

± 1.3 ml. per 100 ml.) is probably near the lowestvaluethatcanbe reached since it is made up

primarilyof bloodfrom working muscle. But the

drop in oxygen content of blood from the brachial

veinofthestationaryarmisalmost as large (4.5 ± 1.2 ml. per 100 ml.) and is in keeping with the re-port of Donald, Bishop, and Wade (24) to the

effect

that during moderate exercise the oxygen content of venous blood from a resting extremity falls as low as that from one whichis exercising. These findings, along with their observation that

the volume of a resting arm is decreased during leg exercise, are most readily explained on the

basisof

physiological

shunting of blood awayfrom resting tissue during exercise. Suchblood as con-tinues to perfuse resting tissue may be so small in amountthatit must surrender as muchoxygen,

relatively speaking, as blood

perfusing

exercising

muscle. In any case, oxygen content of femoral and brachial venous blood during heavy exercise

draws very close, as exercise

proceeds,

to the

cal-culated mixedvenous oxygen.content

(5.0

ml. per 100

ml.).

The moststrikingfeature oftheblood gas

stud-ies is the relative constancy of oxygen

tension,

both arterial and venous,

during heavy

exercise (Table V). In contrast, a decrease in arterial oxygen tension during low

grades

of exercise has

been

reported

by other observers

(26-28).

Blount,

McCord,

and Anderson

(29),

however,

found nochange inarterial oxygen tension

during

similarexercise loads.

Hickam, Pryor, Page,

and Atwell

(30),

and Asmussen and Nielsen

(31)

noted a slight drop

(about

2 per

cent)

in arterial oxygen saturation

during heavy

exercise but the first group

(30)

felt that the

phenomenon

might

be due topH

changes

andnotto a decreasein

(9)

terial oxygentension. Our resultssupportthe

lat-ter view (Figure 2). Oxygen tension of brachial

and femoral venous blood also tended to remain at or near the resting value even though oxygen content and per cent saturation are both markedly depressed during heavy work. Teleologically, it would seem that an adequate level of oxygen ten-sion is vigorously protected so that an effective pressure gradient from capillary to tissue cell can be maintained. The decrease in pH and the in-crease in carbon dioxide tension during exercise combine to cause a temporary shift of the oxygen dissociation curve to the right in the systemic

capillary, but thepHchanges alone were not large enough to explain the magnitude of the shift as actually observed (Figure 3). The carbon diox-ide tension of brachial venous blood increased from a meanvalueof45mm. Hg at rest to65mm. Hg at the maximal oxygen intake level of exercise; thecorresponding figures forfemoral venous blood were 50 mm. Hgand 65 mm. Hg. It may be, as

suggested by Roughton (32), that increasein car-bondioxide tension tends,independently of change in pH, to cause a shift to the right of the oxygen

dissociation curve. In any case, the shift makes it possible, as blood traverses the capillary, for oxygen to be given up more readily without

dis-turbing the pressure gradient from capillary to cell. Inthepulmonary capillary, the situation re-verses itself, facilitating oxygenation of venous blood.

The maintenanceof arterialoxygen tension

dur-ing heavy exerciseisadequate evidence against the possibility that pulmonary factors, ventilatory or

diffusive, determine maximal oxygenintake. Nor did timed vitalometry, carried out at rest in some of the subjects, provide evidence of impairment ofpulmonary function. Onemightwell go further and hold that a definite decrease in arterial oxy-gen tension during heavy exercise, farfrom being

a normal finding, is actually an indication of ab-normal response to exercise.

In Table VI are presented the data on six nor-mal men whose oxygen intake leveled or actually declined atworkloadsbeyondthatproducing maxi-mal intake. It is noted that as oxygen intake

de-clined, the cardiac output also declined. The AV oxygen difference, however, appears to have increased still further. The possibility that the

organism

can increase AV oxygen

difference even after cardiac output and oxy-gen intake have leveled off or are declining is open to several interpretations. It may mean that car-diac output is the more important determinant but

thedata at hand do not rule out the possibilitythat,

in some subjects, further widening of the AV oxy-gen difference may permit oxygen intake to in-crease even after cardiac output has reached its

peak.

In any event, the fullmeaning of maximal oxy-gen intake is in terms not merely of the ability of

the heart mechanically to propel blood, but also

of theabilityofthe tissuesto extract oxygen from

blood perfusing them. It seems unnecessary,

therefore,toinvokethe view thatanimportant

de-terminant of maximal intake is the inability of

peripheral arteries and arterioles to take up all

the blood pumped out by the vigorously

con-tracting leftventricle (2, 16). The evidencenow

available suggests that maximal oxygen intake is ameasure ofcardiaccapacity and the ability to in-creasethe AVoxygen difference, notoftheability of thevascular bed to accommodate left ventricu-lar output. With this in mind, the test may

in-deed have clinical usefulness in evaluating

pa-tients withcertain types of cardiovascular disease.

CONCLUSIONS

1. The maximal oxygen intake is dependent on

both cardiac output and AV oxygen difference. When the oxygen intake increased

by 9.5

times

from rest tothe workload producing the maximal figure, thecorresponding figurefor cardiac output was 4.3 and for AV oxygen difference 2.2. The widening of the AV oxygen difference was due

principallyto diminution in mixed venous oxygen content.

2. There was no significant change in arterial

oxygen tension from rest to heavy work; the slight decrease in oxygen saturation that was

observed can be

explained by

the pH change of the blood and the

resulting

shift in the oxygen

dissociation curve.

3. The venous oxygen tension showed no

sig-nificant change, even

though

the venous oxygen

content and saturation fell

appreciably.

The shift in the oxygen saturation curve to the

right

was

due, in part, to decrease inpH, and

_(possibly)

to

(10)

The end result of the phenomenon is to maintain an adequate oxygen tension gradient from capil-lary to cell.

4. In ascertaining the physiological meaning of the maximal oxygen intake, the relative impor-tance of cardiac capacity and increase in AV

oxygen difference must be determined. It is

probable that in the normal individual the ability to increase cardiac output is the more important

ofthetwofactors.

ACKNOWLEDGMENT

The authors are grateful for the technical assistance of Bernard Williams, B.S.

REFERENCES

1. Hill, A. V. Muscular Activity. Baltimore, The Williams and Wilkins Co., 1926.

2. Taylor, H. L., Buskirk, E., and Henschel, A. Maxi-mal oxygen intake as an objective measure of cardio-respiratory performance. J. appl. Physiol. 1955, 8, 73.

3. Robinson, S. Experimental studies of physical fit-ness in relation to age. Arbeitsphysiologie 1938, 10, 251.

4. Donald, K. W., Bishop, J. M., Cumming, G., and Wade, 0. L. The effectof exerciseonthe cardiac output and circulatory dynamics of normal sub-jects. Clin. Sci. 1955, 14, 37.

5. Dill, D. B., and F6lling, A. Studies in muscular ac-tivity. II. A nomographic description of expired air. J. Physiol. 1928, 66, 133.

6. Brozek, J., and Alexander, H. A note on the esti-mation of the components of variation in a two-waytable. Amer. J. Psychol. 1947, 60, 629.

7. Sproule, B. J., Miller, W. F., Cushing, I. E., and Chapman, C. B. An improved polarographic method for measuring oxygen tension in whole blood. J. appl. Physiol. 1957, 11, 365.

8. Singer, R. B., and Hastings, A. B. An improved clinical method for the estimation of disturbances of the acid-base balance of human blood. Medi-cine 1948, 27, 223.

9. Chapman, C. B., and Fraser, R. S. Studies on the effect of exercise on cardiovascular function. I. Cardiac output and mean circulation time. Cir-culation1954, 9, 57.

10. Hamilton, W. F., Moore, J. W., Kinsman, J. M., and Spurling, R. G. Studies on the circulation. IV. Further analysis of the injection method, and of changes in hemodynamics under physiological and pathological conditions. Amer. J. Physiol. 1932, 99, 534.

11. Scholander, P. F. Analyzer for accurate estimation of respiratory gases in one-half cubic centimeter samples. J. biol. Chem. 1947, 167, 235.

12. Riley, R. L., Proemmel, D. D., and Franke, R. E. A direct method for determination of oxygen and carbon dioxide tensions in blood. J. biol. Chem. 1945, 161, 621.

13. Bainbridge, F. A. The Physiology of Muscular Exercise, 3rd ed., rewritten by A. V. Bock and D. B. Dill. London, Longmans, Green & Co., 1931.

14. Christensen, E. H., and H6gberg, P. Physiology of skiing. Arbeitsphysiologie 1950, 14, 292. 15. Buskirk, E., and Taylor, H. L. Maximal oxygen

intake and its relation to body composition, with special reference to chronic physical activity and obesity. J. appl. Physiol. 1957, 11, 72.

16. Astrand, P.-O. Experimental Studies of Physical Working Capacity in Relation to Sex and Age. Copenhagen, Ejnar Munksgaard, 1952.

17. Slonim, N. B., Gillespie, D. G., and Harold, W. H. Peak oxygen uptake of healthy young men as determined byatreadmill method. _J.appl. Physiol. 1957, 10, 401.

18. Robinson, S. Metabolic adaptations to exhausting work as affected by training (abstract). Amer. J. Physiol. 1941, 133, 428.

19. Wyndham, C. H., and Ward,J. S. An assessment of the exercise capacity of cardiac patients. Cir-culation 1957, 16, 384.

20. Asmussen, E., and Nielsen, M. The cardiac output in rest and work determined simultaneously by theacetylene and the dye injectionmethods. Acta physiol. scand. 1952, 27, 217.

21. Freedman, M. E., Snider, G. L., Brostoff, P., Kimel-blot, S., and Katz, L. N. Effects of training on

response of cardiac output to muscular exercise in athletes. J. appl. Physiol. 1955, 8, 37.

22. Christensen, E. H. Beitrage zur Physiologie schwerer korperlicher Arbeit. V. Mitteilung: Minutenvolumen und Schlagvolumen des Herzens wahrend schwerer k6rperlicher Arbeit. Arbeits-physiologie 1931, 3, 470.

23. Barratt-Boyes, B. G., and Wood, E. H.

Hemody-namic response of healthy subjects to exercise in the supine position while breathing oxygen. J.

appl. Physiol. 1957, 11, 129.

24. Donald, K. W., Bishop, J. M., and Wade, 0. L. A study of minute to minute changes of arterio-venous oxygen content difference, oxygen uptake,

and cardiac output and rate of achievement of a

steady state during exercise in rheumatic heart disease. J. clin. Invest. 1954, 33, 1146.

25. Asmussen, E., and Nielsen, M. Cardiac output dur-ing muscular work and its regulation. Physiol.

Rev. 1955, 35, 778.

26. Lilienthal, J. L., Jr., Riley, R. L., Proemmel, D. D.,

and Franke, R. E. An experimental analysis in man of the oxygen pressure gradient from alveo-lar air to arterial blood during rest and exercise at sea level and at altitude. Amer. J. Physiol.

1946, 147, 199.

(11)

27. Filley, G. F., Gregoire, F., and Wright, G. W. Alveolar and arterial oxygen tensions and the

significance of the alveolar-arterial oxygen ten-sion difference in normal men. J. clin. Invest.

1954, 33, 517.

28. Suskind, M., Bruce, R. A., McDowell, M. E., Yu, P. N. G., and Lovejoy, F. W. Normal varia-tions in end-tidal air and arterial blood carbon dioxide and oxygen tensions during moderate

ex-ercise. J. appl. Physiol. 1950, 3, 282.

29. Blount, S. G., Jr., McCord, M. C., and Anderson, L. L. The alveolar-arterial oxygen pressure

gradient in mitral stenosis. J. clin. Invest. 1952, 31, 840.

30. Hickam, J. B., Pryor, W. W., Page, E. B., and At-well, R. J. Respiratory regulation during

exer-cise inunconditioned subjects. J. clin. Invest. 1951, 30, 503.

31. Asmussen, E., and Nielsen, M. Studies onthe

regu-lation of respiration in heavy work. Acta physiol. scand. 1946, 12, 171.

32. Roughton, F. J. W. Respiratory functions of blood in Handbook of Respiratory Physiology, W. M. Boothby, Ed. Project No. 21-2301-0003, USAF School of Aviation Medicine, 1954.

33. Dill, D. B. Life, Heat, and Altitude. Physiological Effects of Hot Climates and Great Heights. Cam-bridge, Harvard University Press, 1938, p. 13.